I was asked to enter in some kind of best Web Analytics project competition at the end of last month and felt actually unconfortable with the questions asked.
My first reflex was to think about the fact that I couldn’t give out any results from my clients in order to protect them and their business model. My second thoughts emanated once I started reading the questions asked in the form I had to fill in.
I’m sharing here my comments to the organiser of the competition I sent him a month ago:
The thing is that the questions in the document actually reflect what is, and I consider to be and I’m not the only one, wrong with the current perception of Web Analytics within our Benelux market and I doubt this is an isolated matter.
On the one side, this deviated view is mainly supported by the vendors and on the other hand also by the companies that actually decide to step seriously, according to them, into the Web Analytics Journey.
The vendors are trying to make the market believe that it’s all about the best product. Cf. all the latest features released, IPO’s, and very, very bad presentations about successful projects made by them at the different events I’ve attended over Europe. The vendors are of course in for the quick buck as they remain mainly American based companies (except for IndexTools, Instadia and Nedstat amongst others), driven by sales targets on the short term. I haven’t seen yet the notion of maintenance or renewal to be taken into account for their sales targets. Therefore, it’s often a question at each quarter to actually meet those sales targets. That the integrating partner afterwards is up for the job or even the vendor understands internal process or business issues is totally not taken into account. Thing is that installing the product and getting the first basic settings running is actually only the prelude to a bunch of processes that need to be set in place within a company for it to take advantage of this sometimes huge investment that’s being made. But the gains can be immense, if done correctly.
And then you’ve got the companies, the clients. Thy have a problem related to accountability and responsibility. A Web Analytics project is often initiated by IT (or at least IT support is needed and we al know that Marcom & IT are not always best friends). So we go through a classical IT process: product evaluation. When doing that, you miss out on half of the processes that need to be taken into account and most of all, might engage into a path that can turn out to be rather costly for your organisation. Here’s why. Imagine a company decides to engage into the path of WA. Web Analytics means:
1. Measuring anything that generates log files or application related data to directly take decision or
2. Enhance information about the transactions done with your company.
The first one can be related to a public site and the measure of a CMS with email, SEO, traffic acquisition campaigns, RSS feeds, blogs and possible what are called Web 2.0 features and some other stuff. The second reason for engaging into WA is quite close to CRM actually: enhance data or integrate data with other data sources.
Then you also have to:
A. Push the information
B. Analyse it (that’s usually where you have no one left)
C. Adapt your communication to the conclusions (ha! That’s always the big joke: which CMS can actually do A/B or MTV? Who has time and money to do it?) and
d. Continue building your online (or application based) strategy.
Before you’re engaging into such a huge, enormous long term, evolving, change management type of project, it might be interesting to stand still at the following questions:
1. Which applications are used by which departments/business units?
2. Who is responsible for them and who is the technological partner? (can have huge implications on budget as we need to adapt the code quite often to get the data: it’s not often as simple as the vendors tell you)
3. How do these applications affect strategy, turn-over, margins?
4. How do these applications link to the departments/business units objectives? (Believe me, some business units have contradictory objectives related to the same page!)
5. What are success factors or ROI implications of these technologies? How much time do we spend supporting, adapting it vs. how much does it bring us; today and in the future?
6. Once these applications have been prioritised, what would be KPIs?
7. Which products might be best suited to answer our requirements today and in the future?
As you see, it’s only at point 7 that you start talking about a product. And as product pricing can depend upon page views, servers, data integration, accuracy of measures, and a lot of many more things, you might hit it wrong.
WebTrends’ Warehouse in version 8 is a great product. Does anyone know that the pricing depends on the number of “events” recorded in the Warehouse?
I doubt it!
To maybe shed some light on a possible solution, we usually favour to get implicated in step 1, through informal meetings, workshops and books we order and send directly to our customers.
Web Analytics is a Journey, not a destination and should certianly not be viewed as a hype.
Agree, disagree, hate me? let me know. My tantrum is that only fools don’t change their minds (il n’y a que les cons qui ne changent pas d’avis) and I remain open for debate as apparently it makes the world spins of what I’ve been told 😉
Were there’s a discussion, there’s hope for the future, so I’m told or as Jose would say “Talk to me”.
P.S. If you’re a vendor, indulge me. I guess it’s my jewish-duch frankness that needs to speak 😉
P.S.S. René writing: I just wanted to add that it’s funny that Aurélie was already speaking about processes before the Washington’s eMetrics Summit (;-) Eric)